1. Field of the Invention
The present invention relates to a video processing apparatus, a video processing method, and a recording medium for detecting a moving object in a video.
2. Description of the Related Art
Conventionally, as a method of detecting a moving object in a movie, a method of detecting a moving object based on a difference between a reference image and a current image is known.
As a method of detecting an object having a certain feature from a movie, a method of comparing an object in a movie with a predetermined pattern, and detecting an object having a high correlation with the pattern from the movie is known.
Japanese Patent Application Laid-Open No. 2011-166243 discloses a monitoring system that detects a moving object from a movie captured by an image capturing apparatus based on a difference between a reference image and a current image, and among a plurality of detected moving bodies, that further detects a moving object satisfying predetermined conditions regarding a size, a moving speed, or the like.
Japanese Patent Application Laid-Open No. 2009-211311 discloses an image processing apparatus that performs detection using a plurality of local feature amounts extracted from a collation pattern upon detection of a human body from an image by collating the collation pattern with a whole image.
According to the conventional method in which only a method of detecting a moving object (referred to as moving object detection, hereinafter) is used, a moving object that is not an actual target of detection may be detected.
According to the conventional method in which only a method of detecting an object having a high correlation with the predetermined pattern (referred to as object detection, hereinafter) is used, a moving object cannot be detected.
In order to solve these problems, moving object detection and object detection are preferably used together, but the following problem arises to realize such a method.
That is, when a moving object having a certain shape exists in a movie, a region in which the moving object is detected by the moving object detection and a region in which the moving object is detected by the object detection do not necessarily match with each other. Therefore, a certain moving object cannot be accurately detected by collating the result of the moving object detection and the result of the object detection.
For example, in the moving object detection, only a moving part of a certain moving object in a movie may be detected.